[Bugs] [Bug 1151384] Rebalance fails to complete - stale file handles after 202, 908 files

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1151384

alex.smith at hp.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
              Flags|needinfo?(alex.smith at hp.com |
                   |)                           |



--- Comment #3 from alex.smith at hp.com ---
Hi Raghavendra, thanks for the response. The file always exists and I can
always see it during/after the rebalance.

$ ll
/data1/gluster/las/data2/videos/cf/f6/cff65c574d724d373af7f6d90b265b78/1-7d3ec80abe0a304cb6c30d65590f0713
total 4372
-rw-r--r-- 2 2131121 2131121 396680 Apr 23 21:26 video_ep_240k-13.ts
-rw-r--r-- 2 2131121 2131121 383144 Apr 23 21:26 video_ep_240k-27.ts
-rw-r--r-- 2 2131121 2131121 374872 Apr 23 21:26 video_ep_240k-29.ts
-rw-r--r-- 2 2131121 2131121 334640 Apr 23 21:26 video_ep_240k-37.ts
-rw-r--r-- 2 2131121 2131121 602352 Apr 23 21:26 video_ep_400k-01.ts
-rw-r--r-- 2 2131121 2131121 652360 Apr 23 21:26 video_ep_400k-03.ts
-rw-r--r-- 2 2131121 2131121 589568 Apr 23 21:26 video_ep_400k-24.ts
-rw-r--r-- 2 2131121 2131121 540500 Apr 23 21:26 video_ep_400k-29.ts
-rw-r--r-- 2 2131121 2131121 586748 Apr 23 21:26 video_ep_400k-32.ts

This happens on all bricks in this volume, once it gets to 202,908 files
scanned it says it can't find the 202,909th despite its existence.

As you can see from the logs it always happens immediately after the stale file
handle warnings so I presumed they were related.

It would be good to have the rebalance not exit (as complete anyway) after
seeing these errors particularly because as far as I can see they should not be
errors!

Alex.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
Unsubscribe from this bug https://bugzilla.redhat.com/token.cgi?t=IIZRMeF6vC&a=cc_unsubscribe